 Welcome back, everyone. If you're just joining us, welcome as well. Thank you for joining us again for a virtual meeting on the Extremely Low Probability of Rupture Probabilistic Fraction Mechanics Code, or XLPR. This is on running the simulation and retrieving results. My name is Matthew Homiak, and I'm NRC's lead in the Office of Nuclear Regulatory Research for the XLPR program. My counterpart from the Electric Power Research Institute, Craig Harrington, is with us as well. The purpose of today's meeting is to help new XLPR users get started using the code by showing how to run a simulation and retrieve results. The meeting today is a continuation of our XLPR Technical Seminar Series. If you've attended all our meetings so far, do you go ahead and just let us know by sending us a message in the chat here? In the first seminar, we reviewed the models in XLPR. In the second seminar, we showed how to set up the inputs directly using Input Set, File, and the SIM Editor. As part of that seminar, we included a demonstration on how to set up inputs for Challenge Problem 1 in your training manual. In today's seminar, we will show you how to run an XLPR simulation and look at the results. In the demonstrations today, we'll show you how to complete Challenge Problem 1 by writing the inputs that we've prepared in the last seminar. We invite you to follow along with us today. Ideally, using an Input Set, you prepare it for Challenge Problem 1. Or you could use the default Input Set in your XLPR main folder. In our final seminar next week, we'll cover a range of advanced methods, and this will essentially be an extension of the prior seminars on the inputs and today's on running the simulation. We are recording all these seminars for later viewing, so don't worry if you missed any details. You can view those by going to YouTube.com and searching for XLPR and looking for our logo. The video for the first seminar is posted, and I think we'll have the video for the second seminar posted by the end of this week. Here's our agenda for today. We're currently in the introduction and opening remarks portion. And we'll review the file structure again briefly, show you how to run the pre-processor. We'll show you how to set up sampling options and how to control Ron using GoldSim. Then we'll show you how to look at the results and interrogate any errors that you may have in the simulation. After that, we'll show you quickly how to use GoldSim to navigate through the XLPR computational framework. We'll have a break and then we'll open up the floor for a good portion of the latter part of the meeting for any questions that you may have. This is an NRC Category 3 public meeting, and that means the public is invited to participate by providing comments and asking questions throughout. We're using the WebEx platform to deliver the meeting. This is how we plan to take your questions. You can submit them at any time using the Q&A feature, and shown on the slide here is a short review on how you can display the Q&A feature depending on whether you're using WebEx in your internet browser or through the desktop client. At designated points, we may also invite you to ask questions verbally. If you'd like to ask questions then, please use the raise hand feature. We'll call on you and unmute your line. Today, our main presenter is Dr. Cedric Salabare from the Engineering Mechanics Corporation Columbus under contract to the NRC, and I would probably consider him to be our chief expert on XLPR, so we're all on capable hands today. And backing him up is our usual team of experts. That includes Craig and myself, Marcus Burkhart from Dominion Engineering, Nathan Glunt from the Apple staff, and Marge Erickson from Phoenix Engineering Associates. Giovanni Facco from the NRC staff is helping out again as our WebEx host. If you're having any difficulties with WebEx setup, you can just reach out to him as the host directly in the chat feature. Right now, Giovanni should be sharing a short poll with you all. And this is just to help us get a better feel on your experience with XLPR and to help the presentation team here for the content. We'll leave this open for a little while. If you wouldn't mind filling it out, we appreciate it. We'll take a look at the results together a little bit later. Craig, would you like to add anything? I'd get unmuted. I appreciate everyone joining today in the NRC hosting these meetings. I know there are some people still working through challenges, I think maybe with the end-user license agreement to access XLPR and certainly if there are issues that I can help with in that regard, please contact me. Otherwise, thanks for joining, and I'll turn it back to you, Matt. Okay. Thanks, Craig, and thanks again, everyone, for being here today. Okay. With that, I would like to turn it over to Cedric to start our presentation. Thank you, Matt. So as Matt said this time, what we'll try to do is give you an overview of how to run the code. And doing this, we have a set of slides, but what we plan to do is also make a demonstration when we do the running. And as you can see now, I'm sharing my screen. So what I will do is taking some pause from the presentation and implement everything we say, and we invite you, for all of you who have the code, to try on your site to do it. We will do it for challenge problem one, but if you have any input, XL input side, it will work the same way. The result will be different, but all the rest will be exactly the same. So let's go back to the presentation. And for this presentation, as resource, we recommend the user manual and mostly chapter 3. In chapter 3, you will find section 332, which talks about the sample size and how to set up the random seed, section 34, which focuses on running the code. Then section 352 deals with error inspection. And you will find in chapter 5 all the list of potential error codes. And finally, section 353 deals with the results. Another source that's not available yet that will be available in the future is xlprtrfw. This is a computational framework development testing and analysis document. OK, so we will start with the file structure and how to navigate through xlpr. So we discussed about this last time. Once you receive the zip file, you will have several folders. Some of these folders are for information or database. Here we are interested with two folders. The first one is a DL folder that contains all the dynamic links libraries, which are the modules for xlpr. And then the folder called xlpr main, which contain the framework. So it goes in player file, which is a .tsp of the full version, the full version, .tsm, as well as the input xl5, which we mentioned and discussed during the previous seminar, and an xl add-in. So this file is really the one you will work with. And you can create copies of this folder for your different case. As long as the copies are in the same level as the DL folder, the Gaussian file can be run and you can generate results. Once xlpr has been run, the Gaussian file will have all the results in it, so you can copy anywhere and you can access from any location to see the result. As long as you don't run it, you can copy anywhere. One thing about the file name. The xl input file name is called xlpr-2.1 input set of xlsx. If this name cannot be changed, Gaussian requires this name to read the input. However, the Gaussian name itself can be changed, whether it's a full version or the player version, you can change it. So in theory, you can also have multiple Gaussian files in the same folder. They will read the same xl file, but you can have multiple files. And this can be useful if you want to make repletage runs. By repletage runs, we say, we mean, well, you could run the same simulation with different random seed, and then if you have a simulation of 10,000 realizations, you can do it five times and then you have 50,000 realizations. Of course, the DL folder name and the DLM name cannot be changed. And for the same reason, for the xl file, you need the correct name to be called. So that's my step. Now, if you want to run xlpr, you have about six steps. The first one is to set up your program. This was the purpose of the previous seminar. So it's all the option you can set in the input set file directly or using the GUIs that will write into xl. The second step, and I will go through it just after this, is to run the preprocessor. We have two preprocessors you can run in xl here. Once you have done this, you can set up the Gaussian parameter, which are mostly the sampling option and some output settings. And then you can run Gaussian at the first step. Once you run Gaussian, you save. It's really important to save, and I will mention this a little later. And finally, once you have your Gaussian run done and saved, you can save and extract the results. So we'll go through all steps, but the first one that was covered during the previous seminar. So let's start with step number two, which is running the preprocessor. So one question is, why do we want, why do we have preprocessor in XRPR? The reason is to improve run types. There are calculations such as leak rate and stress intensity factor for transient and take time. So we have one preprocessor for each. So to calculate leak rate as a function of crack opening displacement and crack length, we have LiPo, which was developed by Oakridge. It's then for leak analysis of pipeline coverage. And for the stress intensity factor and other information for transient, so if you consider fatigue, we have Tiffany, which stands for thermal stress intensity factor for any equivalent story, it was developed by SA. What these preprocessors do is they create lookup table as a function of the input we mentioned. And GoldSim will just linearly interpolate from the stable during simulation, which is a lot faster than calling the code, adding the calculation at every time step or many calculation at every time step. The purpose of the preprocessor is just speed up XRPR. The preprocessor are accessible via the Excel ID. And so this is the second file, Excel file you have in the Excel PR main, which is called Excel PR-2.1 preprocessor.xll. This ID works only with a 32-bit version of Excel. If you try to use it with a 64-bit version of Excel, you will have an error right now. So this Excel ID has been developed without the digital signature. So when you start it by double clicking on this element, you will first have a Microsoft security notice saying that there's no digital signature, so it can be called DLL. So what you have to do is to enable it and you will have to enable every time you use it. So you have two buttons and you can click on the lower left button, which say, enable this ID for this session only. Once you click this, you will have access to the ID. Now, what's possible is some company security protocol block any macro coming, so they will be disabled. In this case, you will have to check with your IT department to enable this ID. The second thing is, if you have the 64-bit version of Excel, or if you don't know, you may end up to have this message that you see on the bottom left here. It will say that the file format is not recognized and may be corrected. If you press yes or no, it will just not run the ID. Now, let's suppose that you are at the correct portion of Excel and you can run the ID. Once you activate it, you will have an information message on the file location. So this pre-processor must be run where the Excel inputs and the goals in the file are located, because it will open the Excel file and check that everything is here and it also needs to have access to the ID. So once you press OK over this option, you will have a ribbon and then you will be able to run the pre-processor. So here, I will make my first stop from the presentation and show you how to set up everything we have stopped up to now. So let's go back to the screen sharing. So here is a window containing all the data from the zip file. So we are on the XLPRv2.1, I have XLPR main, I have the DLF. So the first thing I will do in this example is make a copy of the XLPR main and I will rename it Challenge Problem 1, because that's the challenge problem we can run. The second thing I will do is copy what we have presented two weeks ago, which is the new input set with the changes and paste it in Challenge Problem 1. And while you don't have to do it, I will delete this repo and tfamil, which are the pre-processor folder, just to show you how they are created every time we call the pre-processor. So here it is, I have my two goals in file, I have the input file and I have the XLR. So the XLR is a file I want to run, so I will double click on it. It will load Excel, so it takes a few times, a few seconds to load everything from Excel. And then you have the message from the Microsoft Excel security, where you want to enable this add-in. When you do this, you have the message and then you will have on the top a new menu in the main reader. So this is called XLPR pre-processing. If you click on this, you have this set, this four set of inputs. So here, because the pre-processor takes some time, I will go ahead from the presentation and press on this, I want to run tfami to have a new transient, even if our case does not include 30. And I will press the start button in the new window to start the execution of tfami. The pre-processor will read all the information from the Excel file and starts to copy it here. And once it starts to run, you will have a bar here that increase in a percentage that will appear to tell you how far you are in finishing to run the pre-processor. So let's move back now, let's move back to the presentation. So as we are saying, we have activated the Excel add-in and now we have access to a new menu. On this menu, you have four buttons. The first button is to inspect the database. So it's look at the table you have created and compare them immediately. We won't cover this in this presentation. What we want to focus is on the screen next button. So this button will run tfami and lipo. We have two buttons for tfami. And the reason is tfami depends on some of properties for the pipe, so pipe geometry. And when you have mitigation, some of the mitigation since it may overlaid may impact the pipe geometry. And so you may need to rerun tfami to have the correct stress intensity factor when you have a different geometry. The first button is for pre-mitigation. The second button is for post-mitigation. And finally, lipo is the last button where you can calculate this great look at table. The next question you may have is, when do you want to run this pre-processor? Oh, sorry. There's no slide before that, which is quite what I have just done. Once you click on the button, you have a new window whether it's for tfami or lipo. If you press the start button, it will start to do the calculation. And once it's completed, you will have a message confirming that the calculation are complete. All the inputs used by tfami and all the output are also saved in the text file with the date and time appended. So it's a unique text file, so you can check which input we have used and which output we have used. Once you have this message, you can close the information box, you can close the tfami, and you have done the calculation. So going back to what I was saying, when do you need to run this pre-processor? But the first question is, what happens if you don't run this pre-processor? Well, if you don't run, Corsim will try to read the last text file generated. If you have this folder, tfami and lipo, and already populated with text file, Corsim will read this text file and just use this data. Corsim does not care whether you have just run or if you have run them two weeks ago. If this file did not exist, Corsim can still run and use what was initially in the lookup table created in the file. So whatever is saved in the lookup table and read because last time will be used. However, Corsim will warn you in a warning message just to let you know that this file didn't exist, so it uses the latest version you had for the data. Now, when do you want to re-run this? The first question is, do you model fatigue? If you don't model fatigue, you don't need to run tfami because these are the stress intensity factor for the training. If you run, if you include fatigue, you need to run tfami at least a pre-mitigation. If you have mitigation, then you will also need to run tfami post-mitigation. If you don't have mitigation, you won't need to run. For lipo, it depends if you have a change in temperature range, pressure range, or pipe geometry. By default, lipo is defined for wide range of temperature, pressure, and pipe geometry to avoid having to re-run it, but if you change, you may have to re-run lipo. In the XR file, and was mentioned during the last seminar, so is the color coding. The color coding that's reported here on the right part of this presentation is from the property tab. So it's a property that RNXL are used by both lipo and tfami. It's highlighted in yellow. If it's used by tfami only, it's highlighted in purple. And if it's used by lipo, it's highlighted in green. This information helps you to know that if you change this value, you may have to re-run tfami, or you may have to re-run lipo. Now, in terms of tips, what we would recommend is if you are unsure, it's safer to re-run the pre-processor, because the date will coincide with the run date and it avoid any confusion. The pre-processor takes some time to run, but usually they still run in 15, 10 minutes, depending on the complexity. And if it's done, you don't have to re-run them, so if it's done once, you can use them. Another point is, lipo and tfami generate new folder with the name lipo and tfami to ask the table. You can delete this folder. They will be recreated if needed. What we recommend if you re-run is to delete this folder and re-run everything so you can confirm that the latest value are used. And once you have turned this, and your pre-processor are ready, you're ready to look at col-sync file and look at the col-sync setting dashboard. So this is when you can run the code. If you have the pro version of col-sync, you will open the .csn file. If you have the player version of col-sync, you will open the .csn file. Whatever you can do with the player file, you will be able to do it with the pro version of col-sync, and you will have more functionalities. So for this demonstration, I will use a player file to show you that everything we present can be done with a player file, but everything can be applied also to the pro version of col-sync. No matter what, when you open any of them, you will start with the col-sync dashboard, which is displayed here in the presentation. I will focus on the first part, which has a tab here. The first one, which is highlighted with a green box, is the sampling approach. These are the sampling options that you cannot set up in Excel and you will have to set up in DOSI. Some of these options are for the epistemic loop, that's left button here. Some are for the aleatory loop, that's the right button here. Whatever you change in the sampling will be reflected in the display below for all the sampling options, for some of the sampling options, like the sample size, whether you use important sampling or not, whether you use adaptive sampling or not, the discretization if you use dpd and so on. This is just for information purposes, to check that this is the run you want to make, you cannot change this, it's just a display. And once you have done this, you have set up your sample the way you want, and if you have continued with the display, you can run the cut. You have two buttons that are like the blue box here. The first one is to refresh all inputs. So this is where once you have done all your changes, you want to be sure everything is okay, you can replace the input. The second one is to run the model, not set. If you run the model by default, Excel PR will refresh all inputs. So whatever has been changed and last, will be taken into account as soon as you press run the Excel PR model. So now we will look in more detail what you can do for the sampling approach. So if you click on the first see upper left button on the epistemic, which is also the outer loop, you will have the option on the sample size. So you have three main options that you will use most of the time. This is the sample size, the use of LHS or not, and if you want to repeat the sampling sequence. So the sample size is pretty straightforward. If you're doing a probabilistic analysis, how many samples do you want to do? Now that's the epistemic sample size. So if you have an epistemic, the outer loop sample size, and an aleatory sample size in a loop, the total number of samples is a multiplication of two. So for instance, you have five epistemic and 10 aleatory, the total number of samples will be 50. So here you set up only the epistemic sample size. The second main option is whether you want to use 19 hypercubes sampling or not LHS. If this box is checked and you use LHS, if it's unchecked then you don't use LHS. Not also that if you check the box, you will have another menu here where you can use a traditional LHS that sample randomly into each strata or you have the midpoint LHS that use the median into each strata. The third main option is the one called repeat sampling sequence. So this is linked to the random number generator. If this box is checked, you control the random seed for the random number generator and you can set it up on the right. That means that your results are reproducible as long as you use the same random seed. By default, it's always checked and you can change the random seed. If you uncheck this box, then the random seed will be generated randomly based on the time of the patient so your results won't be reproducible. Now, as you can see in these windows, there are other options. So we will cover them even though they are not used as much or they are more advanced options. The first one, which is highlighted in green here, is to run a specific realization. So if you're interested for, you run Android epistemic realization that you want to see what's going on in the example of realization side out of Android, you can select it. This is what we recommend for this terminated trend and this will be covered in the next presentation, the advanced method. Another option you see here is to use of weight for the important sample. So here, this is a difference and you can see the framework model we selected in Excel PR. In Excel PR, we consider that you should always add this box check. This box is used when you use important sampling because important sampling tries to focus on one area of your sample space and so because it's over sampling this area, you need to use a lower weight and the under sampling area meets higher weight. So if you don't check this box, important sampling will not work. If you use regular sampling, of course, you don't need a weight, but by default, equal weight will be used so it will work. So here, what we recommend is even if you don't use important sampling, keep this box checked all the time. And for the same reason, both seem the code should include a deterministic simulation which is slightly different from what we have. You could use this in theory, but what we recommend is to uncheck this box Excel PR included on deterministic approach that will be described in the next training and it will be easier to use this approach than trying to use this one. So this box should be always unchecked in the realisation way it should be always checked. On the bottom, you have an information of the size size which we will talk about when we go to the memory management of Excel PR. Now, we have covered this upper left button which is for a epistemic uncertainty. Now we will cover the next button which is for aleatory uncertainty. So because it's a submodel, it's slightly different when you press on it you will be by default on the definition tab so you will have to click on the Monte Carlo tab to have the same control on sample size. So, I liked it in these two blue box these are exactly the same option as the epistemic loop so I won't cover it again. There are a few differences for the aleatory. The first one is on the number of realisation. You can see here there is a lock it means that this value has been locked and instead of having a number you have a reference which sees C017 which refers to variable 107 in the Excel input file. This means that the aleatory sample size is not controlled by costume it's controlled in the Excel file. You have set it up in Excel you don't need to set it up here. The second option that is not included in the epistemic is this one. And to understand this option we have to point out that the epistemic loop in our model is the outer loop and the aleatory loop is the inner loop which means that when we run the code we'll start with epistemic 1 and then we will run all the aleatory. Let's say if your sample size is 20 you will do epistemic 1 and aleatory 1 to 20 then epistemic 2 and aleatory 1 to 20 and so on. Now you have an option here which says use a different random seed for each realization of the parent model the parent model is the epistemic seed. If this option is unchecked it means that you will use the same value again and again for the aleatory each time you change for an epistemic seed. So let me explain this a little more with an example. Let's suppose you have an epistemic sample size of 5 and an aleatory sample size of 10. If this box is unchecked it means that once you run epistemic 1 you will sample 10 value for every aleatory variable but when you move to epistemic 2 the same 10 value will be used again and when you move to epistemic 3 this 10 value will be used again. So even if at the end you do 50 simulations you only sample the aleatory value 10 and you use each of them 5 times. If you check the box for the aleatory different value will be sampled. So once you are on epistemic 1 you will have 10 values. When you go to epistemic 2 you will have 10 different values and so on. So at the end you will sample the aleatory variable 50 times. You will have 50 different values. So here it's really depending on how you want to do your analysis because you want to read it all out. Our recommendation of whoever is to keep this box checked because you have more variability in your aleatory uncertainty. So this cover the different options you have the sampling option you have in your safe. Now we will summarize this with this slide what do you need to set up in Gaussian what do you need to set up in Excel and what is set up in both. So you have a variety of sampling strategy in Excel here. You can use one or two loops and when you use one loop you can use either the aleatory or the epistemic. This will be set up both in Excel and in Gaussian. You can control of course the epistemic sample size and the random seed and this is what we just seen is set up in Gaussian. You can do the same thing with the inner aleatory loop except the sample size is set up in Excel, the random seed is set up in Gaussian. You can apply LHS by just checking the box it's done in Gaussian. You can apply important sampling on selective value. It can be done on either or both loops and you can use discrete probability distribution on either or both loops. All of these can be done together. So you can use LHS with important sampling you can use important sampling with RLHS you can use DPD with important sampling and so on. For the last two they will be set up in Excel. They have been described quickly in the previous presentation on the next presentation we will go a little more detail on how you can run important sampling and how you can use DPD. And finally there is the option we just mentioned about repeating the random data if you want more variability in your inner aleatory loop or if you want to repeat the same thing. And all of these are mixed together so you have a large variety of sampling strategy when you use Excel here. Once you select your strategy it's time to run the code. You can refresh data to be sure that everything is okay but then you're ready to run the code. So we'll do an example with challenge problem one. In the player version of the theme we have three ways to start the run. You can press the button in the dashboard which say run Excel PR. You also have access to the run controller and you can press the play button and finally you have a shortcut with the TS5 on your keyboard. Any of the three will work. Okay, now what we'll do is we'll make a break. We'll go back here to the presentation and go a little more on the run controller but I will demonstrate how you can set up everything we have described and run the code. So a little time to run. We are doing about 200 in this example so it takes about 7 minutes. Then we'll move back to the presentation to talk a little more about the controller. So we have left our example where we were running the preprocessor and as you see Tiffany has finished now so I can close it and close this so Tiffany has been run and we have the calculation. If I go to my folder now, Lipo and Tiffany are created again. You can see Lipo is empty because I didn't run Lipo. If I move to Tiffany now I have every of the tables that are read by Go Sim so it goes from table 51 down to table 84 have been created for the pre-mitigation and they are available which means that Go Sim will be able to read them. Now, as you see when we open XLL it also opens the input set. This is the challenge problem one. We have changed the period one to be 35 years so it's 35 times 12 months and if we go to the properties a little higher we can see that for the second period we get the same pressure but we change the temperature to a distribution. We have set up challenge problem one. If you don't satisfy but still want to test at the same time it's fine. You can use the existing input set to run this. Let me close it now. Nothing has been changed so I can close it. Now we will open the KFI. For this you can double click if you have Go Sim 11.1 or if you have a shortcut you can go here and click on your shortcut for Go Sim. And here I am in the folder for challenge problem one so since I'm already here I don't have to navigate through the structure so I select XLPR 2.1 Go Sim will load everything and then we'll have access to the XLPR model. Now they say it will open to the dashboard. The top part is what we discuss about setting the sampling approach. If I put a piece in it so here you see everything is great. The reason is if you haven't run anything we are in result mode by default Gaussian has been run so we need to go back to the edit mode. We'll cover this later. I click on this button to delete everything. Yes I want to delete everything and I'm going back to the edit mode. It will take a few minutes and we'll see that the message ready is appearing with the controller so now I'm in edit mode and nothing is great so I can work on this. I will change the sample size to 10 and I will say yes I want LHS for the epistemic mode. Now I close it. I click on the addatory button as we said we start so we have to move to the Monte Carlo tab. The number of realizations is set in Excel so I don't change it but I can put for instance the news of LHS. Now this sample size has been updated directly. I could refresh the input and this will disappear to say that we don't consider important sampling but since it will be refreshed anyway I will just run the model. As we say I can press here I can press play on the controller I can press F5 in this case I will just press this way. Now the first thing you will see when you run the model is Colsin will have to download or import all the the external data. So it includes all the DLLs. Colsin will import the DLL put them in memory so it's ready to be used. It includes all the information you put in the Excel file. All the modifications so one needs to be updated so Colsin will read this updates the correct element and say now I have all this correct information. It also includes all the table we mentioned the preprocessing table from Leipol from Tiffany's and so all of this will be uploaded. If they are not available you will have one message at the end we will cover this a little later and then Colsin once everything is downloaded will start running. So downloading all this information takes between 20 and 50 seconds depending on the computer. And then once this is done as I say you go into run mode and the calculation will start. So we will go back to the presentation and discuss about the Colsin run controller and that's how it finishes running. For your information I'm running right now on the laptop. It's five years old laptop. It's still a good laptop but not top of the line so if you run it on a desktop computer or newer laptop it will be probably faster inside the computer. Okay let's move back to the presentation and as we said we were into running Colsin. So we have setup everything and we run Colsin. Now we have this run controller that control everything of running so let's look at this in the internal details. You have three main modes in the run controller. The first one is the edit mode. The second one is the running mode itself. This is the one we have just seen and the third mode is the result mode. It's like once the run is done and the results are available so you don't run but you still have access to all the results that were generated. We will focus on three icons on the bottom part of the controller and these three icons have different meaning depending on which mode you are. The first one on the left is to be ready in edit mode. So it delete all results if there is any result or it's above the simulation and it's free power Colsin to be run. So of course if you're in edit mode already it will be grayed out and this is what you see here in the left of the edit mode. If you're in run mode it means that you want to be in edit mode so it will have both secure and simulation and it will ask do you want to delete the result or not. If you don't delete the result you will go in result mode with whatever you have. If you delete this result you move back to edit mode and if you're in result mode it will delete any result and go back to edit mode. The second button is the one where you want to run. So we have seen this already it looks like a triangle like a play button. If you're in edit mode it just means I want to run the mode if you're in result mode it means that it will delete any information output you have and start to run again. And finally if you're in run mode you're already in run mode this is replaced with a pause button and you can pause and look at where you are in the simulation. Now you have another button which is on the right part of the icon here which is called the menu button. The menu button will give you access to different menu if you have the full version of course seen it's already available on top of your window if you have the player version it's not available so this button is really important. It gives you access to the file option so you can load a new file and you can save or save as. If you have access to navigation if you want to see the inner mechanism of the XRTR model of course if you're in run mode this button will be great because you don't want to save or you don't want to navigate and change something while you're running the code. So let's go into a little more detail of the edit mode and the running mode. So as we say if you're in edit mode you will see it because it's written ready on the right box. It tells you how many realizations you plan to have. The elapsed time is set to 0 because nothing has been run. You have option as a play you can run the simulation and you have the menu option that helps you to save. So if you click on the menu option you will have a rolling menu here and you can go to the file to open save the file. This is pretty straightforward. If you're in run mode now we don't have ready we have run. As soon as you run the model you will have the elapsed time increasing so you can see how much time has spent since the beginning of the simulation. The realization number will indicate which realization is run out of our menu. This case was only one epistemic realization out of one of our informatives not the total realization. You can stop you can pause as we discuss. Now if you want to see the aleatory realization you can do it by hover your mouse on the controller. If you hover your mouse on the controller as you can see down here you will have information window popping up tells you which epistemic realization is run and which aleatory realization is run which time step is run. In this example it was aleatory realization 7. It was just such a big deal. So it can tell you where you are in your simulation. Now once run is finished an important step is saving the run. It's really important. We'll talk about this memory management. Those things save a lot of data and if you try to display them it will require more memory and sometimes you may have the software crashing. It's a fine crash because there is not enough memory. It will open again but nothing will have been saved. So if you have run your case and you crash it then you lose everything and you have to re-run again. So every time the run is performed and is finished it saves the run. So how do you do that? If you're in the player this is where the menu button in the run controller plays a role. You click on it, you go to file and you have an option to save or save it. Not that the shortcut like Ctrl S and Alt plus FA also works. If you're in the full version everything of this is also available. But you have another option on the top here. You have the file menu. You can go and you will have save, save as save a copy and an additional option which is not available in the player file which is called save player file. When you have the full version of your scene you can always save as a player file. Whether you are in edit mode or result mode. Which means you can create a player file as edit mode so people can run with a player and you can make some change. You can also create result version as player so people with a player can look at your result. So it's a really important feature because while the full version of those scenes is not free, the player is free so it means that if you have a full license then we can have people saving results that will be available for anybody who has a player version as long as you can save as player file. Okay. And once the run is finished generally you will have a run log generated. This run log gives you information if anything happened that's not expected during the run. It will also include some information of when the run was performed, when it was finished, along with took and so on. You will have a set of warning or potential error. Usually you will have only warnings. You may want to check this after a while if you notice a warning you may decide not to check. An example of warning is if you have run three runs in pre-processor and you have deleted the text file then go see the warning where I use whatever is saved in GSM or GSP but I was not able to find in text file. Once the run is done you have access to this file and if you press yes you will see it. If you press no and want to look at it this file will be saved in the folder and it's called go see run log.txt So now I will make another break to show you how we can do this and you will have the opportunity to follow on and try it by yourself. So let's go back to sharing the screen. Okay and now you see that I'm done with the run we have the 10 out of 10 realization finished. I'm in result mode because it's finished it took about 9 minutes and 11 seconds to run. And go see purpose me to look at so once. Yes yes and I opened this file the text file I say which is go see run log.txt give me the name of the file run, the starting time, how long it takes and then a set of warning. As you recall we haven't re-run tfn e-post notification and we deleted the result so go see the informer that the file was not found so nothing was imported and the original value was used. As I say an important step is to say before looking at any of the results I go to file save us and we'll save with a different name here as underscore results. It takes generally about 10 to 20 seconds to say depending on the size if you have last size and you need to say 3 gigabytes obviously it takes a little more. If I go here in my folder you can see that now the file is save and the size is about 143 megabytes. I have access to the go see run and I have access also to a narrow log in case there was error in this case we don't have error also that's why it's only one. Okay so this concludes the first part of this presentation we will take a break now the five minute break if you have any question please don't hesitate to use the Q&A to don't hesitate to use the Q&A so we can ask for the question we can answer the question and as usual here are the email you can use in the future if you have any question please send this to bothanasi.gov and api.gov Yes thanks Eric we wanted to go ahead and put up the poll again just because we got a little bit lower response rate than we were expecting so maybe some folks missed that when we put it up in the beginning of the meeting so here it is again mind responding there's just a few questions here we'd appreciate that it's for your benefit at ours and while we're waiting for people to respond there just give everyone a chance to catch up a little bit here and if you have any questions on what we've been showing you so far go ahead and submit something to the Q&A or you can raise your hand as well and we'll see if you can ask your question verbally if you'd like I think today's session is a little bit more mechanical in a way we're just showing you the button presses and things that you need to do to run the code the last session was maybe a little bit more complicated even because we're dealing with all the inputs and things like that I'm not seeing any new questions coming in at this time but if you do have questions go ahead please respond to our poll as well and then turn it back to Cedric and we'll keep going through the presentation here thank you okay so what we have seen so far is how to set up an Excel PR run and how to run up to completion and save the results now the next step is to look at how you look at the results and how you can export them we won't cover everything here some of these will be covered in the advanced section next week so let's come back to the global city dashboard and now we'll focus on the bottom part of the presentation so on the bottom part there's four sets of buttons the left button are here to look at the results and the right button are here to look at the error so the result can be on the top the actual crack and on the bottom the circumcrect in our case we will focus on circumcrect because the only direction we consider for this analysis was circumferential crack in the right part the error dashboard will inform you on potential error or warning that you may have observed well the ghost team may have observed when running the code so one recommendation we do is once you are run your code and once you have saved your code the first thing you should do is open the error dashboard to show that everything has been run as expected if you open the error dashboard you will have information for each of the modules most of the time what you will see is gray square which means that the module was not used for this realization or green check mark the green check mark means that the module perform as expected now sometimes you will see a yellow triangle which is a warning error the warning error may not be serious it may be okay for your analysis but there is something you have to know so in this case you will have to go and click go to the error list and see what's going on and see if the warning is something you have to correct based on your analysis framework you may also have a stop sign which is a fatal error if you have a fatal error you won't usually stop Corsim and arrive to the result Corsim will stop before that and let you know that there is a fatal error so it will be in post-mod and it lets you check what's going on and finally you may have a red flag which means that you have multiple warning or fatal error in the same module so before looking at the results you want to check this if you have only gray square green check mark you're good to go for the result now we have another tip we wanted to mention in terms of running the code sometimes when you have a large sample size and also when you consider many mechanisms such as axial and circumferential such as PWCC and fatigue the code will take some time to run especially if you have a large sample size one frustrating thing is to run the code for 2, 3, 5 hours and discover that you forgot to change something so you have to rerun again 5 hours after so what we recommend and save a lot of time in the long run is to run first with a very simple size just take 1 to 5 minutes confirms that there is no warning or errors check if the result looks reasonable and if so then you bump up to your original sample size and you can run the simulation especially at beginning it's safe time because you have seen that we have more than 500 inputs that you can change and you can put distribution and I think Matt mentioned that there is tens of thousands of cells you can change in this model so it's easy to make an error so it's nice to check first before running a big case and discover after 5 hours that something was not set up as expected ok let's go back to the results now if you click on the result button then you will have a new window a new dashboard that contains several results I won't cover the navigation on the right it's just going back to the global settings or to the different results or error dashboard we'll focus on the rest of this dashboard on the top you have what's called the general result that can say there are both axial and circumferential it can be just probability of having the crack the first crack whether it's axial or circumferential probability of having a leakage whether it's coming from one direction or the other probability of fracture so you have different information that's more generic for the web below that are the direction specific results so these are kind of the same indicator but just for circumferential crack when you're in the circumferential crack result and just for axial crack when you're in the axial crack result so it could be the probability of having a first circumferential crack probability of having leakage due to circumferential crack number of circumferential crack and so forth and below that here highlighted in the red box the crack specific result so these are more for people who are interested in the physical result of crack and look at the evolution of crack what are we brought here are the first crack five crack in terms of occurrence time so crack one means the first crack occurring crack fifth means the fifth crack occurring while you can have both of you set up for an axial crack per direction and you can increase this to a search crack per direction usually with what we realize seeing the first five cracks is plenty so for each of these cracks you can see the type and by type we mean whether it's a surface crack, a swirl crack if it has quelles if you're in a circumferential direction the location of the crack, how much it leaks and the crack property or size, inner length depth, outer length if it's a swirl crack as well as the stress intensity solution in a different direction for all of these you can look at the specific crack for specific realization so you can track the evolution of this crack over time or you can look at summary result like statistics so for all of these but and you have a result element and all of these result elements will work exactly the same way as we will see in the next slide when you press on it you will have the result element opening the result can be either in a chart form or in a table form so we'll look at the chart form first the chart form will give you the evolution of the stress over time so here we have in this example the simulation that ran for 60 years so if you click on chart you will have the visual representation if you click on table you will have a table that looks like Excel in the demonstration I will show you how it looks like once you select this whether you select chart or table you have different representation the first one that chosen by chosen by default is a real realisation so you will see only one realisation when we say only one it's only one epistemic realisation so it's for the outer loop so it means that if you have 20 aleatory you can still have a distribution here the second option is all realisation so here you will see all the epistemic realisation so if you have a sample size of 200 you will see 200 curve now because it's all realisation if we add all output it will be messy so when you're in realisation mode most of this result element will give you a set of information this is what you see in this example we can see the probability of having a first crack probability of having the first leak probability of having rupture different option whether you have in-service inspection or not if you have leak rate detection or not and so on so you can see all these curves if you select all realisation then this that you select which realisation you want to see will not be used but it will be replaced by which output you want to see so realisation here you can change to realisation one to three all realisation you can select which output you want to see probability of having the first crack probability of leak and so on the probability option will be the same thing except it will represent a density where most of this epistemic realisation are and it will show you a density map so same thing you will be able to select which output you want to see the density map and finally you have the statistics that calculate the average or even quantize the median you know 95th percentile overseas data it will work like the realisation only you will see all the output together the other things is you can change the statistics you can see the median you can see the selective part so summary you can select in every result element you can select whether you see a chart on table you can select different representation depending on what you want to see then the next icon is a chart style this is why you can change the X axis you can put log scale and so on you can change the title of the X and Y axis and you have some other option if you are on table mode this allows you to sort the results so you can look at the maximum value and so on and finally there is a button that you can use to export the data which is called the access to result element this pattern will only be available in the pro version I will show you you can still access to the element in the player version it will require a little more steps it's still available but it's kind of a shortcut if you have a pro version so let's look at this access to the result element so as I say in the next slide I will show you how you can do this with the player version and I will demonstrate but once you have access to this element you can see all the outputs that are selected for this element you can put them also on a different scale on the Y axis if you want so you can display them or not you can change the color and the style for each of these an important option is you can export all these results as text file or Excel file by default this will be set to none if you change to text file or Excel file this is a button on the bottom part of it you will have a new tab that will appear at the top of this windows it's called export if you click on export you will have this is here in the down part you will have an option in the bottom part you will have an option to export the data as a text file or as an Excel file an important point even if we haven't done it if you want to put a new name here you will need to do this in edit mode go see by default once you in the result mode that's not that you allow you to change anything both sim want the result to be fixed and everything so everything is fixed so it does not allow you to change also a text file to export the results so if you are in edit mode you don't change the name if you are in result mode you don't change the name and by default the name is blank.txt and blank.xlsx you can still save it you can still change the name once you are in windows it's just like a strange name by default but you will still be able to export and we will show you when I demonstrate this this is as this result mode is directly accessible with the pro version if you have the player version you will need a few extra steps you will need to navigate through the model so you remember this golfing run controller you have this menu button instead of going to file and save the results you will go to navigation and you will go to model root go to model root just go open a new window and on the right side you will have all the results that are displayed you look at the results that have the same name as the one you are interested like here it was comparison with fiction and if you right click on it you will have an option which is property and when you right click on the property you will go to this property for this result element so the same thing as the pro version okay so when we come to the error now I will move back to our little example and show you how to set this up okay so we are here we have run the mode we have the result we have saved it as a different name and we have the result mode as I say the first thing we recommend is to click on the error dashboard to check that everything is fine in zero dashboard we see only gray squares and green check mark so we are good so we can go to the top now on the top you have some navigation button you can go back to global setting with the middle one you can go to the actual error with the right one the left one will bring you directly to the circumferential result so that's the one we want to go I click on it and you can see what we have seen previously which is a circumferential result with a general result on the top it's a specific direction specific result in the middle and the track specific results so as in the example I will run the detection effect as a result and here you see what as a result for realization epistemic realization 1 as we have 20 addatory we still have a distribution and not just an occurrence with probability for epistemic realization 1 if I change here I can see realization 2 3 where nothing happened 4, 5, 6 and so on I can change here the display I can decide to see also realization if I press here now I can look at the occurrence of crack I can look at the occurrence of leak I can look at the occurrence of rupture this should include 10 result because we have 10 realization I can see this as a density function using the probability button the probability button will give you a density where the darker are closer to the median and the lighter are further from the median and it show kind of different context it give you an idea of where most of the result are with respect to the epistemic and finally you can look at the statistics where you can change from mean, median and so on not that let me blow at this if you move over the curve it will tell you which curve it correspond and it will give you the X and Y locations so if you want to see what's going on at 30 years you may not need to go to the table and look at here and you see that the probability at 30 years was around 14% of having the crack the next button we mentioned was the chart style so if I press on the chart style a new menu will open where I can change the different properties let's go to the Y axis and let's suppose that I want to present this in logarithmic state I press logarithmic and I apply R will give crossing will calculate automatically which has a bound you don't necessarily want to go to 10 to minus 8 here I can put 10 to minus 3 and apply and now I will have the result in large scale between 10 to minus 3 and 1 and once again I can move my mouse to see what the different result are once this is done you can close and you have access to your result now as we mentioned this is a chart you can see it or see as a table data the table data will look kind of as an Excel spreadsheet and you can select any value and check what's going on if you click on result on the top you will select the full set of data if you click on a specific column you will select this column if you click on specific row which is a time set you will select everything that you can do with Excel like control shift down or control shift up or control down up will work with code theme so you can take it you can copy and paste with control c, control v and copy into Excel or any of the tools you want as I mentioned the chart type becomes sorting for the column so you can sort a column or you can sort a row when you select this actually a row I'm not sure it's working maybe only for the column so you can sort your data and look at the evolution of the data so this is hard to look at the result we can also export if you have a version this icon is great so you don't have access if you have the pro version of the thing you can click on it directly if you're in the player you will have to go to the ghosting controller take on the menu go to navigation go to model root in the right part of the module root I moved with this and here we have all this data we were looking at comparison detection I have property and here is my property menu right now I have only one tab if I move here and I say I want to export as a text file now I have the export tab I cannot change the name but I can still export if I press it will export it says the export is completed if I go to my folder now you see I have this blank.txt file which I can change if I want and I can open it to have all the results exported so this talks over the reading of results now we'll move back to the presentation and we'll look at in detail more detail what to do when you have an error the most common error will appear in ghosting where you have a message when you run the code it changes to error you will know which epistemic and editorial simulation gives you an error you can reset the simulation if you know what the code is and start again or you can check what's going on also covered in this slide is a post mod even if you don't have an error you can always post the simulation and go into realization epistemic realization by realization to see what's going on well if you have an error most of the time what will happen is ghosting will give you an error message tells you which mode is affected and where the location is so here we'll do an example where we will change the crack rows to be negative which is not accepted by example you don't want the negative crack rows where actually instead of cracking it starts to grow again what you will have is it will tell you this realization has a problem so we stop the simulation this element and the message is the module the actual crack row model is affected so in this case except here it does not accept negative crack rows even if we don't run actual crack actual crack check on this and we'll say oh you have a negative crack row I have to stop and let you know now there is a difference between ghosting error and excel PR error these are excel PR error we have set up a method just to check if something is wrong and we have a way to track this but it's not a ghosting error or ghosting we'll just interrupt and say this element has me to interrupt so when you have this you can press continue and you will have the simulation continuing or you can press pause and pause will link you to the interrupt element which may not be informative so what you need to do is go back to the global dashboard go back to the error and then you will be able to track what's going on one point that's kind of important is even if you're in the player version when you're in this error debugging mode you will suddenly have access on the top to all disruption with a file and so on and different menu options so you will navigate through it to go back to the root so what's happening is we'll move back to the error dashboard this time we don't have only real check mark and gray box we don't have a stop button so we'll go to the error list if we go to the error list we'll see that the error flag is written as 106 and 106 means that in the list of error it means that the power load constant you can read this here the power load constant is out of French that's to be positive and we put a negative number so this helps you to track which error so let's take a break from the presentation here move back to the screen and let's create this error in our example okay just to be on the same side I will open this file so the results I've run are saved in a different file so I just open the original file and I will open also I will open also the excel input file seems that I close by mistake the folder where we have everything so then we go back to excel pr to perform to other folder and to change problem one so now I open again the original file and I will open the input set and create this error in the quad rows right I want to move back to edit mode so I can run the part again so the quad rows factor is defined in the web folder I go to this power load constant alpha and instead of 2 times minus 12 I change it to minus 3 I will be covering the advanced but already the input file give you which branch should be used and if you don't use the correct branch it will highlight in red so you should have the information that something is not correct so even if it's not correct we will save it and we will run go scene again so now we press play we will have this loading import thing of all this data as we had previously but this time once the import is finished the code will start to run and that's the first epistemic and first aleatory realization when the module discovers that this is not a value will have to stop indicating there is something wrong you cannot run this simulation so the importing data is sometimes a little longer we don't know exactly why ok so we have the last time running that's strange because it should give you an error here for some reason so here you have a very good example of the Marfillo we have tested this example several times before that and every time it created an error so I don't know what's going on here unless I haven't changed the correct file this is a correct file this is challenge problem one let me check that I have a correct file ok I will try another time but if it doesn't work we will stop here and we will go with a different kind of error ok well let's give it just 30 more seconds if it's not working we will move back to the next set of error which is the memory management and if you try this on your own normally you have reached a case where you have an error by changing this and the code has stopped ok for some reason this does not work while I tested this morning and it was working so let's move back to the presentation and go into the memory error so here what we use for the framework is Corsi 11.1 it's a 32 bit program by default 32 bit program are limited to 3 gigabytes of data when you run even if you run on a 64 bit machine and OS if the amount of data is required is larger then Corsi will crash because there's no memory installed then it will start to re-run so that's why it's important every time a run is complete to be sure that you don't go beyond this the biggest indicator for the memory requirement is in the Monte Carlo tab in the inner loop the outer loop just tabs the results just save outside of the main model so it's not very informative in the inner loop you will see that if you change if you have a larger sample size this number the purple number down will increase you an indication in our experience if you try to go beyond 3 to 4,000th realization in the inner loop you may start to see some error you can expand by using both the epistemic and the inner loop you can do with a current model up to 8,000th realization for instance, 80 epistemic times under the outer loop from testing it seems that the formula is every time you run go scene you need at least the minimum 80 megabytes of initial memory plus a third of a megabyte per realization so each three set of realization add one megabyte which means that the maximum size with the original model is around 8,000th realization now they seem low but there are ways to do more and the next seminar will present ways to increase this limit larger sample size for your information we have ranked cases where you have 100,000th realization by limiting and by setting to some specific and in this graph you can see the size of the go scene file once it's saved once you have 20 realization 100 realization 200 realization 400 realization once you pass the beginning ok, so we have covered how to know how to run go scene how to look at the results how to handle with error despite this strange behavior now I would like to cover something a little aside which is how to navigate through the go scene model we have approached it slightly at one time when we wanted to navigate because one button was not available in the player file these will be useful when you move into the more advanced method it's also useful when you want to look what's going on in the model and better understand all the modular interactive shows if you have a new version of go scene you will have access to the model directly with a button on the top you have two buttons the first one moves to the root container the other one goes to the main root container which you press one or twice you will go to the root if you have the player version you don't have access to the button so you will need to go to the navigation button so the main button navigation and go to model root model and this is what we have done when we wanted to have access to the result element now if you move to the root model you will have this it's an object oriented program so you have objects and depending on which kind of object it is you have a different presentation so that's the go scene graphics representation you will have also these kind of boxes which have containers you can think of them as folder where we group the different things each of the container has a plus button on the top if you press the plus button you will go inside the container so to navigate you will just go in and out of the container and if you want to go out of the container you will have the buttons here on the top left that will be available it's directly available it uses the full version if you use the player version you will need to go to the model root first now what I will show you is how to access what we have presented in the previous seminar which is a time loop for the second print software so we will go to the main model we will click on it it will bring us here here you will have all this module that all these containers that take the information from Excel and so on calculate additional data and send this to the XRPR model so we will click on the XRPR model in the XRPR models you have the first container for the initiation then you have the circumferential loop some post processing and some error checking and results check so we will go to the subcrack and then you will have this circumferential crack loop where you have the time loop with the initiation calculation of the case solution and growth because it's a circumferential crack the transition, the stability and as an example we will go to the crack growth and look at the PLL that calls this growth module so navigating through it may still be hard because there is a lot many layers in XRPR you can have access to the flow of information with any element if you right click on an element you will have access to function of affect for both of them you will see all the information from this element for instance if you right click on the DLL element it will tell you all the elements that are sent so you can make the calculation like for crack growth you may need to know the initial crack you may need to have a case solution and all of these will be done so if you right click and go to function of you will see all the elements that are affecting crack growth affect is the opposite it looks at downstream information you will see the elements that the DLL affect once it have calculated crack growth where does it send the information whether you are in function of or affect you can press this plus button to see now which element are affecting this one and so on so on you can go as deep as you want and if you click on one element you will be sent to this element so you can see what is going on for instance if I go to the crack growth DLL and I click on this it will send me to this element so I can look what is going on with this element we can change a container we can turn from one container to the other so it may be easy to be lost in this case so that's why you have these two blue arrow buttons that bring you to the previous container you visited and to the forward container so let me demonstrate this quickly with an example here so here it is we are in result mode and we want to see what is going on in this element so it's available both in result and edit mode so I go to navigation and go to model so now I have access to this previous and next container I will go to the main model as I say I will go to excerpt your model and then to subtract and this is my time look I can go to subtract growth and then I have access to my DLL if I right click on the DLL I can have access to the function so I can see which element affects this DLL I can go as deep as I want so I know that this alpha GM to affects the controller and I can see that it's this element if I click on the element I jump to this element directly whichever it is situated in this section so this is this element if I move here back and forth I may not know where I am so I'm lost but if I press here I will go back to this element and I will go back to the DLL so these are kind of the basic manipulation tools you can have it navigates to the model to learn more about the model it's not necessary to do this when you run the code it will be more useful you will see in the advanced method that will be presented next week when you want to change things when you want to save different output when you want to do other things navigating through the model is a necessity in this case and let me go back to the presentation now and this is the conclusion of this presentation we are done so we have presented you how to run the model and look at the results now we are available for any questions you may ask and we'll be happy to answer thank you thank you Cedric I was able to reproduce your error there Cedric let's go to have some questions if you have any questions from the presentation go ahead and submit those in the Q&A in the meantime or raise your hand and we'll put up your line for you files crossed so Cedric I'll I think we have lost Matt can you hear me I can hear you Cedric okay so as Matt said to discuss anything you want don't hesitate to raise your hand or to ask questions on anything and we'll be happy to go back to the presentation and to make a test for you Craig do you want to add something we don't know where Matt disappeared to I'm sure he'll be back momentarily I don't see any other questions at this time heard part of what you said Matt yeah can you hear me now Cedric I think I have your example fixed okay I'm the presenter now hopefully the audio issue is cleared up I'm going to share my screen here now I don't think that's going to work I apologize for that I'm having connection issues here yeah we apologize for that we are just aware sometimes that by sharing we are taking some memory some strengths like we don't know I think it's kind of a memory sharing issue but for instance when you we can see your screen now okay all right maybe we can get this to work then so what Cedric had basically shown was that the power law constant on the crack growth rate on the on the well tab was negative you see that in the input set it shows up as red that gives you initial warning but if you were to go ahead and run that input set which I'll do right here and we should see that error that Cedric was trying to show so we'll just give it a moment here and again if you have go ahead and send them in I'm supposed to get ready to run we should see that error pop up here just a second there it is then I think Cedric you were wanting to show show pause right so you could investigate the error yes okay close okay so now you have access on the top to the different parts if you go up to the top next to the blue arrow but first here I cut this one with the one you pick twice not this one the next one twice we go to control dashboard which is in the center just below XORP2.1 and now we can go to error dashboard axial crack since it will be an axial crack first and we see the stop on crack growth rate for ID we go to the error list and we see that the first crack is the error flag is 106 and if you read along the description of error the 106 is a power local instead out of front of the DT okay good yeah so that was what we were trying to show earlier so we're glad we were able to get that resolved I'm going to stop sharing that now thank you so I see we have a question about the limitation about there is limitation for the number of pallets we do for realization due to a problem although it's extremely low probability so I will do two answer on this the first one is in the advanced technique we'll see that it's possible to do larger sample size with some change especially if you're interested only with a mean value and you can you reduce the number of time-dependent results you save you should be able to run 10,000 or even more 20-50,000 per innovation which means that you can be in the range of 10 to the minus 4 10 to the minus 5 markers will present also some techniques that reduced by saving even if you calculate that every time step you can save the result not at every time step and so you can also gain on this the second thing is when we talk about extremely low probability it's often when we consider inspection and recreate detection and for instant inspection will be probability that we apply to it saying well you have one chance out of 100 to detect it or one chance out of or 99 chance out of 100 to detect it so it's only one out of 1% of chance not to detect it so when you have this factor applied even if you have 10,000 realization when you apply the factor saying well it's 10,000 realization but you have only 1% of chance to miss this and to create a problem you have this factor applied so instead of being 10 to minus 4 you go to 10 to minus 5 or 10 to minus 6 probability so this is a case where the probability will go down just because the likelihood of not detecting it is lower and finally and that's something we have used you can do replicates which means that if you have a case with 10,000 realization you can do replicates by changing the random seed and make 10 case of this and suddenly you have 100,000 realization in the analysis we have performed for a constant check between nursing a pre for instance we have case where we run 15,000 realization 7 times that give us 100,000 realization which means that we can detect around 10 to minus 5 so I hope this answers your question about this extremely low probability if you have if you need more detail please don't hesitate to add another question and let us know what you want to know specifically I don't see any other questions for now and we are approaching the end of this presentation so I will mark well except here is a complex code and we know that we cannot cover everything and there are many things that you learn when you use a code so we understand that right now you may not have questions in the proper later so please if you have any questions or if you block somewhere when running except here don't hesitate to send us an email and we will try to answer as early as possible to any other question whether you know it's on setting up the input or running to code looks like maybe we lost Matt again I don't think so we have one last slide yes let's I think Matt is yours now yeah I say somehow I'm the host and I'm not sure how that happened okay can you move to the last slide please Matt technical difficulties maybe if I accept being the host I can move to the last slide oh yeah is this the one you were looking for now that was for the Q&A let's go keep closing remarks so so we would just like to remind everyone that we do have one last seminar next Wednesday same time slide on August 5 and we will be covering sort of an eclectic collection of advanced methods several things that Cedric made reference to today in just other trips tricks and tips that may be useful to you as you attempt to use XLPR and learn some of the ways to accomplish the kinds of analyses that it's capable of and so we will get information out of how to join that webinar and look forward to your participation there as Matt and Cedric have mentioned all of these are being recorded and we will be making that information available to everyone how to access them but if you go to YouTube and search on XLPR I believe you'll be able to find these recordings as they're listed but we'll have other other pathways to help you get to this information as well Matt did you manage to get back with us? Yeah I think so can you hear me? Yes Okay I apologize for that I think on the NRC side here we've just had some issues both Giovanni and I let's see so I will try to pick it up here and close this out to the end of the meeting here see Craig you let's click you reviewed our plans here for the next session that's next week I don't really have much else I don't see any outstanding questions or anybody raising their hand either so let's go ahead and close this out of course if you have any questions or feedback you can submit that to us at xlprnrc.gov and xlpr at every.com and we want to thank you again for your participation today and if you'd like as in the other seminars we if you'd like to provide feedback on NRC meetings generally you can do that through the NRC public website meetings section you just find this meeting you can fill out a form there we hope to see you next week and otherwise or email us that's for me when the videos are posted I think that's it thanks again everyone for participating and I hope to see you soon and we'll adjourn thank you